• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1825-1838
• /
• 2016

The detection of skin pigment is crucial in the diagnosis of skin diseases and in the evaluation of medical cosmetics and hairdressing. Accuracy in the detection is a basis for the prompt cure of skin diseases. This study presents a method to recognize and measure human skin pigment using Hemoglobin-Melanin (HM) coordinate. The proposed method extracts the skin area through a Gaussian skin-color model estimated from statistical analysis and decomposes the skin area into two pigments of hemoglobin and melanin using an Independent Component Analysis (ICA) algorithm. Then, we divide the two-dimensional (2D) HM coordinate into rectangular bins and compute the location histograms of hemoglobin and melanin for all the bins. We label the skin pigment of hemoglobin, melanin, and normal skin on all bins according to the Bayesian classifier. These bin-based HM projective histograms can quantify the skin pigment and compute the standard deviation on the total quantification of skin pigments surrounding normal skin. We tested our scheme using images taken under different illumination conditions. Several cosmetic coverings were used to test the performance of the proposed method. The experimental results show that the proposed method can detect skin pigments with more accuracy and evaluate cosmetic covering effects more effectively than conventional methods.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.1
• /
• pp.363-372
• /
• 2017

This paper presents a novel face recognition algorithm based on the deep convolution neural network and key point detection jointed local binary pattern methodology to enhance the accuracy of face recognition. We firstly propose the modified face key feature point location detection method to enhance the traditional localization algorithm to better pre-process the original face images. We put forward the grey information and the color information with combination of a composite model of local information. Then, we optimize the multi-layer network structure deep learning algorithm using the Fisher criterion as reference to adjust the network structure more accurately. Furthermore, we modify the local binary pattern texture description operator and combine it with the neural network to overcome drawbacks that deep neural network could not learn to face image and the local characteristics. Simulation results demonstrate that the proposed algorithm obtains stronger robustness and feasibility compared with the other state-of-the-art algorithms. The proposed algorithm also provides the novel paradigm for the application of deep learning in the field of face recognition which sets the milestone for further research.

• Journal of The Korean Astronomical Society
• /
• v.51 no.4
• /
• pp.73-88
• /
• 2018

We present a sample of 54 disk galaxies which have well developed extraplanar structures. We selected them using visual inspections from the color images of the Sloan Digital Sky Survey. Since the sizes of the extraplanar structures are comparable to the disks, they are considered as prominent stellar halos rather than large bulges. A single $S{\acute{e}}rsic$ profile fitted to the surface brightness along the minor-axis of the disk shows a luminosity excess in the central regions for the majority of sample galaxies. This central excess is considered to be caused by the central bulge component. The mean $S{\acute{e}}rsic$ index of the single component model is $1.1{\pm}0.9$. A double $S{\acute{e}}rsic$ profile model that employs n = 1 for the inner region, and varying n for the outer region, provides a better fit than the single $S{\acute{e}}rsic$ profile model. For a small fraction of galaxies, a $S{\acute{e}}rsic$ profile fitted with n = 4 for the inner region gives similar results. There is a weak tendency of increasing n with increasing luminosity and central velocity dispersion, but there is no dependence on the local background density.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 2017.04a
• /
• pp.154-154
• /
• 2017

Root is an important organ of plant that typically lies below the surface of the soil. Root surface determines the ability of plants to absorb nutrient and water from the surrounding soil. This study describes an application of image processing and computer vision which was implemented for rapid measurement of seed germination such as root length, surface area, average diameter, branching points of roots. A CCD camera was used to obtain RGB image of seed germination which have been planted by wet paper in a humidity chamber. Temperature was controlled at approximately 250C and 90% relative humidity. Pre-processing techniques such as color space, binarized image by customized threshold, removal noise, dilation, skeleton method were applied to the obtained images for root segmentation. The various morphological parameters of roots were estimated from a root skeleton image with the accuracy of 95% and the speed of within 10 seconds. These results demonstrated the high potential of computer vision technique for the measurement of seed germination.

• ALGAE
• /
• v.34 no.4
• /
• pp.315-326
• /
• 2019

Northeastward drifts of massive Sargassum patches were observed in the East China Sea (ECS) and Yellow Sea (YS) by the Geostationary Ocean Color Imager (GOCI) in May 2017. Coverage of the brown macroalgae patches was the largest ever recorded in the ECS and YS. Three-dimensional circulation modeling and Lagrangian particle tracking simulations were conducted to reproduce drifting trajectories of the macroalgae patches. The trajectories of the macroalgae patches were controlled by winds as well as surface currents. A windage (leeway) factor of 1% was chosen based on sensitivity simulations. Southerly winds in May 2017 contributed to farther northward intrusion of the brown macroalgae into the YS. Although satellite observation and numerical modeling have their own limitations and associated uncertainties, the two methods can be combined to find the best estimate of Sargassum patch trajectories. When satellites were unable to capture all patches because of clouds and sea fog in the ECS and YS, the Lagrangian particle tracking model helped to track and restore the missing patches in satellite images. This study suggests that satellite monitoring and numerical modeling are complementary to ensure accurate tracking of macroalgae patches in the ECS and YS.

• Journal of The Korean Astronomical Society
• /
• v.30 no.1
• /
• pp.83-94
• /
• 1997

We have identified the candidates for the primordial galaxies in the process of formation in the Hubble Deep Field (hereafter HDF). In order to select these objects we have removed objects brighter than 29-th magnitude in the HDF images and smoothed the maps with the Gaussian filters with the FWHM of 0.8' and 4' to obtain the difference maps. This has enabled us to find. very faint diffuse structures close to the sky level. Peaks are identified in the difference map for each of three HDF chips with three filters (F450W, F606W, and F814W). They have the apparent AB magnitudes typically between 29 and 31. The objects identified in different wavelengths filters have a strong cross-correlations. The correlation lengths are about 0.8'. This means that an object found in one filter can be also found as a peak within 0.8' separation in another filter, thus telling the reality of the identified objects. This angular scale is also the size of the primordial galaxies which have strong color fluctuations on their surfaces. Their large-scale distribution quite resembles that of nearby galaxies, supporting the idea that these objects are ancestors of the present bright galaxies forming at statistically high density regions. Inspections on individual objects show that these primordial galaxy candidates have tiny multiple glares embedded in diffuse backgrounds. Their radial light distributions are quite different from that of nearby bright galaxies. We may be now looking at the epoch of galaxy formation.

• Journal of the Korea Fashion and Costume Design Association
• /
• v.11 no.2
• /
• pp.85-98
• /
• 2009

The modern trend in costumes, influenced by postmodernism, is to use of various patterns and images borrowed from diverse cultures of many ethnic groups. The Yo tribe studied in this paper is miner ethnic group in China, whose traditional costume is very splendid and modern. In addition, its embroidery pattern has a high artistic value in that its shapes are diverse and splendid and each one has its own peculiar elegance. As for the research method, I examined the Yo tribe's history, culture, traditional costumes and design patterns through related books, research papers, internet sites, and etc. The results of the paper are as follows. The Yo tribe's costumes consist of a jacket, trousers or a skirt, an apron and a belt. Although the color of the costumes is all black, there are splendid embroidery decorations with the colors of red, orange, yellow, green and white on the chest or shoulder part of a jacket, the adjusting lines, cuffs, or a part of a trousers and aprons. The types of the patterns represented in the Yo tribe's traditional costumes are related to nature, ancestor worship, ethnic legends, history, religion, and agricultural lives. The method by which the Yo tribe expressed on their costumes is a "peach-blossom" technique, which uses cross-shaped embroidery with wrap and woof threads. Because it is not apt to express delicate and detailed patterns, the Yo tribe's patterns tend to show abstract and geometrical forms.

• Journal of KIISE:Software and Applications
• /
• v.30 no.5_6
• /
• pp.430-443
• /
• 2003

The objective of this paper is to evaluate most useful feature vector space using the angle, velocity and location features from gesture trajectory which extracted hand regions from consecutive input images and track them by connecting their positions. For this purpose, the gesture tracking algorithm using color and motion information is developed. The recognition module is a HMM model to adaptive time various data. The proposed algorithm was applied to a database containing 4,800 alphabetical handwriting gestures of 20 persons who was asked to draw his/her handwriting gestures five times for each of the 48 characters.

• Journal of Biomedical Engineering Research
• /
• v.28 no.2
• /
• pp.265-270
• /
• 2007

Polarization imaging systems have been widely used to selectively characterize skin lesions. Nevertheless, current systems are used in single-mode due to the limitations of a fixed polarization mode and a single-working distance of light source, in which uniform light distribution is achieved on target area. To address such limitations, we developed a variable polarization imaging system based on multi-working distance of light source for various clinical diagnoses. In this study, we characterize the imaging system and present experiment results demonstrating its clinical usefulness. The imaging system consists of a CCD color camera, linear polarization filters, and a single-layered LED ring light source which provides uniform light distribution at multi-working distances. The first polarizer was placed on the light source and the second polarizer placed on objective lens provides continuous linear polarization angle from $0^{\circ}\;to\;90^{\circ}$. The clinical efficacy of the imaging system was investigated by acquiring and analyzing clinical images of skin wrinkle and dental plaque. With the experiments, we verified the potential usefulness of the imaging system for other clinical applications.

• Journal of Digital Convergence
• /
• v.7 no.2
• /
• pp.93-102
• /
• 2009

Since video is composed of unstructured data with massive storage and linear forms, it is essential to conduct various research studies to provide the required contents for users who are accustomed to dealing with standardized data such as documents and images. Previous studies have shown the occurrence of undetected and false detected shots. This thesis suggested shot verification and video retrieval system using visual rhythm to reduce these kinds of errors. First, the system suggested in this paper is designed to detect the parts easily and quickly, which are assumed as shot boundaries, just by changing the visual rhythm without playing the image. Therefore, this enables to delete the false detected shot and to generate the unidentified shot and key frame. The following are the summaries of the research results of this study. Second, during the retrieving process, a thumbnail and keyword method of inquiry is possible and the user is able to put some more priorities on one part than the other between the color and shape. As a result, the corresponding shot or scene is displayed. However, in the case of not finding the preferred shot, the key picture frame of similar shot is supplied and can be used in the further inquiry of the next scene.